Lecture 20 Instrumental variable estimation

Lecture 20 Instrumental variable estimation - Economics 326...

Info iconThis preview shows pages 1–7. Sign up to view the full content.

View Full Document Right Arrow Icon
Economics 326 Methods of Empirical Research in Economics Lecture 20: Instrumental variable estimation Vadim Marmer University of British Columbia May 5, 2010
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Endogeneity I In the linear regression model, Y i = β 0 + β 1 X i + U i the condition for consistent estimation of β 1 by OLS is that X is exogenous: Cov ( X i , U i ) = 0 . I When Cov ( X i , U i ) 6 = 0 , we say that the regressor X in endogenous. I When the regressor is endogenous, the OLS estimator is inconsistent: ˆ β 1 , n β 1 = 1 n n i = 1 ( X i ¯ X n ) U i 1 n n i = 1 ( X i ¯ X n ) 2 ! p Cov ( X i , U i ) Var ( X i ) 6 = 0 . 1/18
Background image of page 2
Consequences of endogeneity I The causal e/ect of X on Y is not estimated consistently: ˆ β 1 , n ! p β 1 + Cov ( X i , U i ) Var ( X i ) . The e/ect can be over or under estimated depending on the sign of Cov ( X i , U i ) . I 2±18
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Sources of endogeneity There are several possible sources of endogeneity: 1. Omitted explanatory variables. 2. Simultaneity. 3. Errors in variables. All result in regressors correlated with the errors. 3/18
Background image of page 4
Omitted explanatory variables I Suppose that the true model is ln Wage i = β 0 + β 1 Education i + β 2 Ability i + V i , where V i is uncorrelated with Education and Ability . I Since Ability is unobservable, the econometrician regresses ln Wage against Education , and β 2 Ability "goes" into the error part: ln Wage i = β 0 + β 1 Education i + U i , U i = β 2 Ability i + V i . I Education is correlated with Ability : we can expect that Cov ( Education i , Ability i ) > 0 , β 2 > 0, and therefore Cov ( Education i , U i ) > 0 . Thus, OLS will overestimate the return to education. 4/18
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Simultaneity I Consider the following demand-supply system: Demand: Q d = β d 0 + β d 1 P + U d , Supply: Q s = β s 0 + β s 1 P + U s , where: Q d = quantity demanded, Q s = quantity supplied, P =price. I
Background image of page 6
Image of page 7
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 03/24/2012 for the course ECON 326 taught by Professor Whisler during the Spring '10 term at UBC.

Page1 / 19

Lecture 20 Instrumental variable estimation - Economics 326...

This preview shows document pages 1 - 7. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online